Atos Launches Codex, an AI Platform for Sensitive and Regulated Sectors
Through a press release published on October 26, 2026, Atos presented an artificial intelligence offering designed for organizations operating in highly regulated environments and sensitive to data matters. The stated objective is to help them transition more rapidly from experimentation to production deployment, with possible deployment in public cloud, sovereign cloud, on-premises, or in environments isolated from any external network.
Codex, a Platform to Convert AI Analysis into Action
Presented as the culmination of the suite, the Atos Codex platform (for "Center of Operations and Data EXcellence") targets so-called critical environments. It is designed to transform data, expertise, and analyses produced by AI into actions executed within a secure and governed framework. According to the group, the tool connects data, operational context, cybersecurity controls, and AI-based reasoning within a single platform. Codex builds a structured representation of the organization's operational environment (entities, relationships, rules, authorizations, and possible actions), referred to as an operational ontology. Decisions can then be translated into actions according to the governance model and level of human control defined by the user. The platform relies on Atos's experience in managing complex environments and on Eviden's expertise in cybersecurity products and critical systems. Identified targets are critical infrastructure operators, transportation stakeholders, industrial organizations, defense agencies, and public institutions. Philippe Salle, Chief Executive Officer of Atos Group, indicated that Codex should enable these organizations to "transform reliable analyses into operational actions" while maintaining "full control of their data and critical operations".
A Suite Built on Open Source and Models Deployable Locally
The Atos Genesis Intelligence suite brings together data management, AI model development and deployment, as well as their orchestration. Its architecture is based on open source technologies, in order, according to the group, to avoid dependency on a single vendor. The Data Foundation module covers the entire data lifecycle, from ingestion to analysis, with quality, cataloging, and traceability functions. The offering integrates intelligent model routing and cost optimization, presented as levers for more frugal AI usage. It also includes an SLM Factory dedicated to small language models deployable locally, as well as a portal providing access to a marketplace of generative AI agents customizable according to use cases. The suite is to be continuously enriched by the Atos Sovereign Agentic Studios, a global delivery model oriented toward the production deployment of AI agents. Atos claims more than 51,000 employees serving more than 4,500 clients in 54 countries.
